Buried in OpenAI's Own Research: No Correlation Between AI Use and Revenue per Employee
A 69-page OpenAI-co-authored study (with Columbia Business School and Wharton researchers) analyzing enterprise ChatGPT usage from Oct 2025–Mar 2026 finds token usage per employee is not meaningfully associated with revenue per employee once controls are applied. Senior leaders use AI least intensively even as they drive adoption decisions; early-career employees use it most.
Why it matters
Read at Fortune Leaders relying on adoption/usage metrics as a proxy for AI ROI are measuring the wrong thing; the paper suggests real value capture depends on empowering frontline usage rather than top-down mandates.
Save this story, pin it to the top of your feed, or keep a private note on it.
Sign in